Tree-Ring δ¹³C and Intrinsic Water-Use Efficiency Reveal Physiological Responses to Climate Change in Semi-Arid Areas of North China

Abstract

Climate change has had a widespread and profound impact on global temperature and precipitation patterns, especially in semi-arid areas. Plant δ¹³C and iWUE indicate the trade-off between carbon uptake and water loss, which is pivotal for understanding plant responses to climate change. Information about the long-term responses of the physiological and ecological processes of different tree species to climate change is also required. To investigate the impact of different forest stand structures and site conditions on long-term growth and physiological processes of coniferous and broad-leaved trees in the mountainous area of Beijing, we analyzed the tree-ring δ¹³C variation of four tree species (Platycladus orientalis, Pinus tabuliformis, Quercus variabilis, Robinia pseudoacacia) sampled from 64 plots with varying site and stand conditions. We found that the tree-ring δ¹³C of the four tree species varied from each other and was mainly affected by density and slope aspect, followed by slope and age. Both tree-ring δ¹³C and iWUE of the four tree species showed increasing trends over time, mechanistically linked to long-term changes in global CO₂ concentration. This indicates the four native tree species have adapted well to climate change, and the risk of decline is relatively low. The increased iWUE translated into different growth patterns which varied with tree species, site, and stand condition. Different tree species have varying sensitivities to environmental factors. The iWUE of coniferous tree species is more sensitive to climate change than that of broad-leaved tree species, especially to temperature (T), the Standardized Precipitation Evapotranspiration Index (SPEI), and vapor pressure deficit (VPD).

1. Introduction

In the past 100 years, the global average air temperature has increased significantly, and the rate of this increase is accelerating, indicating that the rising trend of air temperature will not weaken in the short term [1,2]. Climate change has had a widespread and profound impact on global temperature and precipitation patterns [3]. The “fertilization effect” caused by the increase in CO₂ concentration, combined with the opposing effect of drought stress inhibiting tree growth, leads to great uncertainty in growth patterns [4,5]. As the main component of the terrestrial ecosystem, understanding the physiological responses of different tree species in this situation is of great significance for predicting tree growth patterns and the process of tree decline in such climate-sensitive areas. Photosynthesis and transpiration in trees are two important ecological processes that are interconected. The increase in atmospheric CO₂ concentration can promote tree growth, but rising temperatures and drought have a negative impact on the forest ecosystem. Under long-term climate change, understanding the response of tree water and carbon processes is of great significance for the integrated management and regulation of regional water and carbon resources. This is especially true in North China, which has become a high-risk area for drought due to the intensification of climate warming [3,6]. Understanding the physiological responses of different tree species to climate change is crucial for understanding and predicting vegetation growth patterns, tree decline processes, and the driving factors in climate sensitive areas.

Carbon isotope effects exist in the process of plants absorbing and utilizing atmospheric CO₂, influenced by photosynthetic pathways (C3, C4, and CAM) and environmental factors [7]. Plant δ¹³C reflects the interactions between various aspects of plant carbohydrate relationships, and can be used more effectively as a comprehensive indicator of whole-plant function compared to gas exchange measurements [8]. It can also reveal how species survive in specific habitats and ensure competitive advantage by regulating their gas exchange processes, resource acquisition, and utilization strategies [9,10]. The dynamic fractionation of ¹³C during plant photosynthesis is influenced by individual differences and environmental factors [7]. Stable isotopes in tree-rings are a valuable tool for studying the physiological and ecological mechanisms induced by drought, and are often used to reconstruct past changes in temperature and precipitation during growing seasons [11,12]. However, their climate response is still not fully understood, particularly in non-extreme mid-latitude environments with diverse ecological conditions. Tree-ring stable carbon isotope parameters reflect the response of trees to climate change over a long period of time. By tracing the physiological and ecological responses of trees during their growth process, and analyzing the differences between different tree species, we can reveal how different tree species respond to climate change [13]. As an important indicator for evaluating tree growth efficiency, intrinsic water-use efficiency (iWUE) directly reflects the impact of global climate change on the ecosystems [14]. The iWUE can be inferred through δ¹³C, which simultaneously records the carbon assimilation and stomatal conductance, both of which are susceptible to environmental conditions [15,16]. In the context of climate change, some forests have declined, while others have adapted to changing environmental conditions. However, the underlying mechanisms remain unclear. The information about the various physical and physiological mechanisms of iWUE and tree-ring δ¹³C change as tree growth progresses is also required.

In order to clarify the long-term physiological and ecological responses of different tree species to climate change, we investigated the impact of different forest stand structures and site conditions on long-term growth and physiological processes of coniferous and broad-leaved trees in semi-arid North China, using tree-ring stable isotopes. Furthermore, we discuss the adaptability of four native tree species to climate change.

2. Methods and Materials

2.1. Study Area

The study area is located in Beijing, China (39°28′–41°05′ N, 115°25′–117°30′ E). This area experiences a temperate continental climate. The average annual precipitation is 500–600 mm, while the average annual evaporation is about 1000 mm. The majority of precipitation occurs in summer, accounting for 64%–69% of the annual total. The average annual temperature is 10–12 °C. The geomorphology of the study area is complex, with vegetation covering more than half of the total area, playing an important role in ecological protection. There are various tree species in the study area. The distribution areas of each species are approximately 86,000 hm² of Platycladus orientalis, 77,000 hm² of Pinus tabuliformis, 53,000 hm² of Quercus variabilis, and 15,000 hm² of Robinia pseudoacacia.

2.2. Sample Plot and Tree-Ring Collection

Two coniferous species, P. orientalis and P. tabulaeformis, and two broad-leaved species, Q. variabilis and R. pseudoacacia, were selected for the study, which are widely distributed naturally in North China and have a long history of growth. These trees have strong adaptability to soil with no strict requirements for texture, and they are tolerant of drought and barren conditions. The environmental factors affecting δ¹³C are categorized into stand factors (including age and density), site factors (including slope and aspect), and climate factors (including precipitation, air temperature, and relative humidity). Due to the thin soil layer in the Beijing mountainous area, and the small altitude variation (400–1000 m) in the main distribution areas of the studied tree species, further subdivision was not performed. Changes in soil water content are mainly influenced by precipitation, slope, and aspect, with precipitation having the most significant effect. Therefore, the response to soil water content is encompassed within the response to slope, aspect, and precipitation, and is thus no longer discussed separately.

Sixteen plot types for each tree species were selected in this study, resulting in a total of 64 sample plots (Figure 1). The specific classifications are shown in

Table 1. Basic information on different plot types in the mountainous area of Beijing. Based on stand density, the plots were classified into low-density plots (density < 1000 plants/ha) and high-density plots (density > 1000 plants/ha). Based on stand age, coniferous forest plots were divided into young forest plots (tree age < 50 a) and middle-aged forest plots (tree age > 50 a). Broad-leaved forest plots were divided into middle-aged forest plots (30 a < tree age < 50 a) and near-mature forest plots (tree age > 50 a). Based on slope, the plots were classified into gentle slope plots (slope < 15°) and steep slope plots (slope > 15°). Based on aspect, the plots were classified into sunny slope plots and shady slope plots.

	Density		Age		Slope		Aspect
	High	Low	Mature	Middle	Steep	Gentle	Sunny	Shadow
S1	√		√		√		√
S2	√		√		√			√
S3	√		√			√	√
S4	√		√			√		√
S5	√			√	√		√
S6	√			√	√			√
S7	√			√		√	√
S8	√			√		√		√
S9		√	√		√		√
S10		√	√		√			√
S11		√	√			√	√
S12		√	√			√		√
S13		√		√	√		√
S14		√		√	√			√
S15		√		√		√	√
S16		√		√		√		√

. Plot surveys and log scaling were carried out on all plots, recording the basic information of each 20 m × 20 m plot and the data of each standing tree. Within each plot, four trees were selected for sampling. Two cores were taken from each tree at breast height (1.3 m) along the east–west and north–south directions, respectively, using a 5 mm increment borer. Except for the damaged tree cores, a total of 512 samples were collected (133 cores for P. orientalis, 137 cores for P. tabulaeformis, 123 cores for Q. variabilis, and 119 cores for R. pseudoacacia). To prevent contamination from other carbon sources, the collected samples were stored in glass tubes.

2.3. Tree-Ring Stable Carbon Isotope Analysis

The core samples were sanded with grain paper of varying frits from 100 to 800 mesh to make the tree-rings more clearly visible for cross-dating. COFECHA (version XP 2007) was then used to verify dating, identify errors in the individual series, and improve the cross-dating accuracy [17]. Prior to the isotope analysis, the first 10 years of each core were removed to avoid “juvenile age effects” on the tree-ring isotopes. Therefore, the tree-rings from 1960 to 2020 were ultimately preserved for isotope analysis. All tree cores from each plot were used for stable carbon isotope analysis after the establishing the tree-ring chronology. The cores were separated into individual rings using a scalpel to cut along the ring lines under a stereomicroscope (40× magnification). Rings from the same year were pooled to form a composite sample representing the plot-aggregated isotopic composition changes [18,19]. Whole wood was used for the δ¹³C analysis in this study to trace temporal trends and responses to climate variables, based on the strong correlation between the carbon sources of whole wood and cellulose [20,21]. Each pooled sample was ground into a powder using a ball mill, sieved through an 80-mesh sieve, and then the milligram wood powder was placed into tin capsules for δ¹³C analyses using mass spectrometry (DELTA V Advantage; Thermo Fisher Scientific Inc., Waltham, MA, USA; precision 0.1‰).

2.4. Carbon Isotope Discrimination and iWUE

The carbon isotope effect in plant photosynthesis can usually be expressed by the following formula [22]:

$${\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}={\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{a}}-\mathrm{a}-\left(\mathrm{b}-\mathrm{a}\right)\times {\displaystyle \frac{{\mathrm{C}}_{\mathrm{i}}}{{\mathrm{C}}_{\mathrm{a}}}}$$

(1)

where δ¹³C_a and δ¹³C_p represent the ratios in atmospheric CO₂ and tree-rings, respectively; a and b represent the isotope fractionation coefficient of the CO₂ diffusion process (4.4‰) [23] and the carbon isotope fractionation coefficient of the photosynthetic carboxylase carboxylation process (27‰) [24]; C_i and C_a are the intercellular and ambient CO₂ concentrations, respectively. The isotope effect can also be expressed by isotope discrimination value (Δ):

$$\mathsf{\Delta }={\displaystyle \frac{{\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{a}}-{\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}}{1+{\mathsf{\delta }}^{13}{\mathrm{C}}_{\mathrm{p}}}}$$

(2)

Therefore, Δ is also related to the ratio of the C_i to C_a mole fraction [24] as:

$$\mathsf{\Delta }=\mathrm{a}+(\mathrm{b}-\mathrm{a})\times {\displaystyle \frac{{\mathrm{C}}_{\mathrm{i}}}{{\mathrm{C}}_{\mathrm{a}}}}$$

(3)

$$\mathrm{iWUE}={\displaystyle \frac{\mathrm{A}}{{\mathrm{G}}_{\mathrm{s}}}}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{a}}-{\mathrm{C}}_{\mathrm{i}}}{1.6}}$$

(4)

iWUE (μmol mol⁻¹) is defined as the ratio of net carbon assimilation (A) to stomatal conductance to water vapor (g_w). The linear relationship between C_i/C_a and Δ can be used to calculate the iWUE [25]:

$$\mathrm{iWUE}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{a}}\left(\mathrm{b}-\mathsf{\Delta }\right)}{1.6\left(\mathrm{b}-\mathrm{a}\right)}}$$

(5)

2.5. Climate Data

Annual meteorological data (precipitation, temperature, and relative humidity) of the study area from 1960 to 2020 were downloaded from the National Meteorological Science Data Center (http://data.cma.cn/, accessed on 12 June 2023). Kendall tests were conducted to confirm the data validity and representativeness of the local climatic trends. The Standardized Precipitation Evapotranspiration Index (SPEI) was calculated using the SPEI package in R 3.5.1 (http://sac.csic.es/spei/, accessed on 5 September 2023) and was used for drought assessment and analysis. Vapor pressure deficit (VPD), an important index reflecting the degree of atmospheric drought, was calculated according to the general formula [16].

2.6. Data Analysis

One-way analysis of variance (ANOVA) was used to test for significant differences in tree-ring δ¹³C between different plots and different tree species. Random forest analysis was used to identify the effects of site conditions on tree-ring δ¹³C by constructing a random forest model incorporating tree-ring δ¹³C and various factors. Regression analysis was used to explore the trends in tree-ring δ¹³C and iWUE. Pearson’s correlations were used to indicate the influence of precipitation, temperature, relative humidity, SPEI, and VPD on tree-ring δ¹³C and iWUE.

3. Results

3.1. Long-Term Tree-Ring δ¹³C of Different Tree Species

The average value of tree-ring δ¹³C for the four tree species ranges from −25.2‰ to −23.2‰, with significant differences among species. The order is as follows: δ¹³C_P.o (−23.963‰~−23.054‰) > δ¹³C_P.t (−24.340‰~−23.737‰) > δ¹³C_Q.v (−24.718‰~−24.203‰) > δ¹³C_R.p (−25.187‰~−24.794‰) (Figure 2). The annual variations in tree-ring δ¹³C of the four tree species between 1960 and 2020, shown in Figure 3, indicate that the average tree-ring δ¹³C of coniferous trees is higher than that of broad-leaved tree species during the study period. Although the growth trend of δ¹³C_P.o has weakened in the past decade, it still shows a significant gradual increase, while the trends for other tree species are relatively gentle. Additionally, δ¹³C_P.o exhibits the largest fluctuation range, with a larger standard error than other tree species, indicating more dispersed average values.

3.2. Comparison of Tree-Ring δ¹³C among Plots and Species

Differences in tree-ring δ¹³C among plots are shown in Figure 4, revealing both similarities and differences among the tree species. The density of plots P1~P8 is relatively high, while the density of plots P9~P16 is relatively low. According to the numerical orders of δ¹³C among different plots for four tree species shown in

Table 2. The numerical orders of δ¹³C among different plots for four tree species.

Tree Species	Numerical Order of Tree-Ring δ13C among Plots
P. orientalis	P1 > P3 > P6 > P5 > P13 > P2 > P9 > P7 > P10 > P11 > P15 > P4 > P8 > P14 > P16 > P12
P. tabulaeformis	P1 > P3 > P6 > P2 > P5 > P9 > P13 > P7 > P11 > P8 > P4 > P10 > P12 > P14 > P15 > P16
Q. variabilis	P1 > P3 > P6 > P5 > P2 > P9 > P7 > P4 > P15 > P10 > P11 > P13 > P8 > P14 > P12 > P16
R. pseudoacacia	P1 > P3 > P6 > P5 > P2 > P9 > P7 > P4 > P15 > P10 > P11 > P13 > P8 > P14 > P12 > P16

, we found that the δ¹³C of high-density sample plots (75%~87.5%) is higher than that of low-density plots, suggesting that density has a positive effect on tree-ring δ¹³C. Plots P1~P4 and P9~P12 are relatively older stands, while plots P5~P8 and P13~P16 are relatively younger stands. The average ranking shows that tree-ring δ¹³C_P.o and δ¹³C_P.t do not exhibit a clear pattern with age, whereas 62.8% of δ¹³C_Q.v and δ¹³C_R.p values are higher in older trees compared to younger ones. Plots P1, P2, P5, P6, P11, P12, P15, and P16 are steep slopes, and the rest are on gentle slopes. There is no obvious regularity shown in the rank of different slopes. Plots P1, P3, P5, P7, P9, P11, P13, and P15 are on sunny slopes, while the rest are on shady slopes. Tree-ring δ¹³C values for 62.5%~75% of the sample plots on sunny slopes are higher than those on shady slopes.

Random forest analysis was used to determine the influence of density, aspect, slope, and age on tree-ring δ¹³C of P. orientalis, P. tabulaeformis, Q. variabilis, and R. pseudoacacia (Figure 5). The results show that the tree-ring δ¹³C of four tree species was mainly affected by density and aspect, followed by slope and age. There are also differences in the importance ranking of these factors for different tree species. For example, tree-ring δ¹³C of P. orientalis is most affected by aspect, while others are most affected by density. In addition, tree age has a negative effect on tree-ring δ¹³C of P. orientalis, P. tabulaeformis, and Q. variabilis, but a positive effect on R. pseudoacacia.

3.3. Relations of Tree-Ring δ¹³C and iWUE with Climate Index

There are significant differences in Δ among the four tree species, with Δ_R.p > Δ_Q.v > Δ_P.t > Δ_P.o (Figure 6). Decreasing trends in Δ were observed over time for stands of all tree species, with approximately downward-slopes of the curves. The intrinsic water-use efficiency (iWUE) of the four tree species, derived from tree-ring δ¹³C measurements (Figure 6), showed similar increasing trends. Both Ci and iWUE show similar increasing trends, with the ranges of Ci being Ci_R.p > Ci_Q.v > Ci_P.t > Ci_P.o and iWUE being iWUE_P.o > iWUE_P.t > iWUE_Q.v > iWUE_R.p. The growth rates of iWUE for the four tree species are similar, with iWUE_P.o having the largest growth rate at 0.88 μmol mol⁻¹ per year.

Meteorological data displayed significant fluctuations in precipitations and rising air temperatures between 1960 and 2020. No significant increasing or decreasing trend was found for VPD, but the VPD anomaly exhibited increasing occurrences of higher VPD. In addition, the frequency of negative SPEI increased. Correlation analysis of the averaged tree-ring δ¹³C from 16 sample plots of four tree species with meteorological factors and drought shows that tree-ring δ¹³C has negative correlations with P and RH, and positive correlations with T, SPEI, and VPD (Figure 7). Overall, the tree-ring δ¹³C of coniferous tree species is more sensitive to climate change than that of broad-leaved tree species. Comparison among the four tree species reveals that δ¹³C_P.o is the most sensitive to climate change, showing a significant positive correlation with T, SPEI, and VPD (p < 0.05), and a significant negative correlation with P (p < 0.05). The sensitivity of the tree-ring δ¹³C responses of the four tree species to P is significantly different, while the sensitivity to T is slightly different. There are no significant differences in the correlations between tree-ring δ¹³C and RH.

Similar results are found in the correlation of the averaged iWUE of the four tree species with meteorological factors and drought, showing that iWUE has negative correlations with P and RH, and positive correlations with T, SPEI, and VPD (Figure 8). It is also found that the iWUE of coniferous tree species is more sensitive to climate change than broad-leaved tree species, especially to T, SPEI, and VPD. The responses of iWUE to P are relatively weak compared to those of tree-ring δ¹³C. There are no significant differences in the correlations between iWUE and P, or between iWUE and RH.

4. Discussion

4.1. Effects of Site Conditions, Stand Factors, Climate Change, and Drought on Tree-Ring δ¹³C

Tree-ring stable isotopes can exhibit robust climate signals [26,27]. However, understanding the influence of various parameters on tree-ring δ¹³C is challenging due to interactions among environmental and stand parameters [28]. The analysis results indicate that tree-ring δ¹³C in four tree species is primarily affected by the density and slope aspect, followed by slope and age. Light and water availability fluctuate in the dynamic field environment. Understory trees face more growth limitations due to drought [29]. The sampling sites differ in soil moisture and light, primarily controlled by slope differences and aspect [11]. This site appeared to be the driest due to the shallow soil, steep slope, and southern exposure. Because drought and shade have contrasting effects on isotope discrimination, it is necessary to distinguish between water and light limitations on tree growth. Our study shows no significant differences in tree-ring δ¹³C between trees growing on steep slopes and those on gentle slopes, suggesting that within the sampling slope range, the physiological processes of the four tree species were not affected by different slopes. There may be two possible reasons for this phenomenon. First, the sampled trees did not distinguish between uphill and downhill slopes. Second, the gentle slope may not have an impact on soil moisture. Compared to slope degree, the impact of slope direction is more significant in this study. CO₂ assimilation is greater on sunny slopes compared to shady slopes [30]. Trees growing on sunny slopes have significantly higher tree-ring δ¹³C than those growing on shady slopes. Similarly, Chen et al. [31] found that plants grown on the sunny slopes were enriched with more ¹³C compared to those on the shady slopes, attributed to the drier and warmer climate on the sunny slopes relative to the shady slopes.

Stand density influences the intensity of interspecific and intraspecific competition in forests, thereby affecting water-use strategies and tree growth [32,33]. However, there is limited understanding of the relationship between stand density, age, droughts, stable carbon discrimination, and tree growth in plantations [34,35]. Tree-ring δ¹³C records signatures influenced by neighborhood effects, where the structure of tree canopies is shaped by stand density and functional traits of neighboring trees [36,37]. In this study, we found that 75% to 87.5% of high-density plots had higher tree-ring δ¹³C than low-density plots. Higher tree densities increase competition for resources (light, water, and CO₂) in the ecosystem. Trees in this forest type enhance water-use efficiency, leading to higher tree-ring δ¹³C values and increased iWUE. Gouveia and Freitas [32] found stand density-dependent differences in leaf carbon isotope discrimination in a water-limited oak forest. They identified an optimal stand density by the comparing of δ¹³C signatures among stands with varying stem densities. Nevertheless, it is still unclear if tree-ring stable isotopes contain age-related trends. In addition, it is important to understand the changes of climate response in tree-ring data of different tree ages, yet there is limited knowledge on the impact of tree age on the climate response of tree-ring stable isotopes. While Gagen et al. [38] observed signs of the juvenile growth effect of δ¹³C in Finnish pine trees, they found no long-term age-related trend in tree-ring stable carbon isotopes. Likewise, Daux et al. [39] did not observe age-related low-frequency behavior in the carbon isotope characteristics of Larix decidua from the French Alps, consistent with findings from live pine and residual pine in northern Sweden [40]. In addition, Young et al. [41] reported that there was no long-term age-related trend in the stable carbon and oxygen isotopes of living pine trees grown in the marine conditions in Northwest Norway. Xu et al. [42] measured tree-ring δ¹³C in three age groups of Schrenk spruce (Picea schrenkiana) trees in northwestern China to determine if age differences in tree-ring δ¹³C could lead to varying climate responses. They found that young trees exhibited a stronger climate response in δ¹³C compared to old trees.

4.2. Tree-Ring δ¹³C and iWUE Reflect Long-Term Tree Responses to Environment

The precipitation pattern and air temperature in North China are changing, manifested by significant fluctuations in precipitation and rising air temperatures. The increased frequency of negative SPEI values indicates the intensification of drought severity. iWUE reflects the balance between the carbon uptake and water loss, influenced by atmospheric CO₂ concentration and water availability. When exposed to long-term water constraints, trees improve their iWUE to reduce transpiration water loss. This is achieved through strict stomatal regulation to maintain the hydraulic function of the xylem, or through stomatal closure due to the loss of hydraulic conductivity. This understanding is crucial for interpreting plant responses to climate change [43,44]. It is important to assess the effects of climate change, such as increasing atmospheric CO₂ concentration and water stress, on carbon sequestration and iWUE [6,45]. A meta-analysis revealed a ~40% increase in tree iWUE globally since 1901, paralleling a ∼34% increase in atmospheric CO₂. However, mean iWUE, and its rate of increase varied across biomes and leaf and wood functional types [46]. We found clear trends over time in tree-ring δ¹³C and iWUE for four native tree species, mechanistically linked to long-term changes in global CO₂ concentration, which is consistent with previous studies [18,47,48,49]. The long-term growth of iWUE indicates an increase in net photosynthesis and/or a decrease in stomatal conductance (and consequently transpiration). The prolonged rise in atmospheric CO₂ concentration is expected to promote both photosynthesis and reduce transpiration [50]. Drought often enhances iWUE, improving the adaptability of trees by enhancing carbon assimilation and reducing water loss. However, worsening drought stress does not always lead to increased iWUE. In the American Southwest, a severe, multidecadal drought has significantly increased plant water-use efficiency and has reached a critical threshold [43]. Currently, a section of the Three-North Shelterbelt Forest in China has experienced decline, particularly in poplar plantations [19,51]. Despite extensive research on the decline of plantations in arid and semi-arid areas, the understanding of the underlying mechanisms remains unclear [52,53]. Integrating tree-ring data and iWUE to elucidate the growth process of trees is crucial for understanding both the mechanisms of decline and the adaptability of trees to future climate change.

Increased iWUE has significantly positive impacts on tree recovery (17%) and resilience (15%), indicating enhanced restoration capabilities [54]. Our study shows that tree-ring δ¹³C and iWUE are still increasing, suggesting that the physiological parameters of the four tree species have not yet peaked. This performance indicates that these native tree species have adapted well to local environmental changes, with a relatively low risk of decline. The increased iWUE has resulted in varying growth patterns depending on tree species, site, and stand conditions [46,55]. Tree species exhibit varying sensitivities to environmental factors. Our study also revealed that coniferous tree species are more sensitive to climate change in terms of iWUE, particularly to T, SPEI, and VPD, compared to broad-leaved species. Both shrub and tree iWUE showed greater sensitivity to variability in atmospheric aridity than to increasing atmospheric CO₂ concentration [44]. Liu et al. [56] determined that temperature and precipitation were the primary factors influencing water-use efficiency in trees growing in seasonal drought areas, based on correlation analyses between needle δ¹³C values and meteorological data. They found that temperature had a stronger impact on water-use efficiency in middle-aged forests. Identifying and selecting tree species with high water-use efficiency is crucial for sustaining forest benefits under changing climate conditions and limited water supply [57].

5. Conclusions

Our study reveals distinct differences in the tree-ring δ¹³C among four native tree species, primarily influenced by density and slope aspect, followed by slope and age. Over time, both tree-ring δ¹³C and iWUE have shown consistent increases, correlating with global CO₂ concentration changes. Different tree species exhibit varying sensitivities to environmental factors. Coniferous species, for instance, demonstrate higher sensitivity to temperature, the Standardized Precipitation Evapotranspiration Index (SPEI), and vapor pressure deficit (VPD) compared to broad-leaved species. Tree-ring δ¹³C and iWUE serve as integrated indicators of environmental variability and the impacts of stand and site conditions on tree physiology and growth. Our findings underscore the adaptive capacity of these native tree species in response to changing environmental conditions. The increasing trends in iWUE highlight their ability to optimize water-use efficiency amidst rising atmospheric CO₂ levels.